Web 2.0 Asia

Read/write web reports that "At the Web 2.0 Summit in San Francisco today, Adobe CEO Bruce Chizen said that the company is working toward shifting all of their apps online".

Programs like Adobe's graphic software require lots of computing power, and it would be a challenge to run those programs entirely off the web, making the servers do all the work. Therefore, it might be possible that what look like pure web programs will actually install snippets of applications, trying to tab into the local computing power.

I think those little program snippets used along with web apps make sense - after all, the computing power of today's desktop PCs is on par with that of supercomputers in the 60's. But in general, I'm not such a big fan of a web service that downloads something onto my PC. Maybe I'm being too tired of Korean internet banking and numerous other services that require downloads of Active X.

Chosun Ilbo, Korea's top newspaper, wrote a piece on Naver today criticizing the service for being too commercialized.

The article gives querying "Chinese" as an example, where a whopping 90+% of the search results are of commercial nature. While Google makes a rather clear distinction between search results and ads, Naver doesn't provide clear visual cues on ads, the article claims. Entering "Chinese" on Naver actually gives a full page load of ads, but these ads are guised as normal web search results as they appear under rather ambiguous names such as "Power Links" and "Plus Pro". I call this strategy shrewd but evil.

Daum is Korea's #2 internet portal site, which also operates under Lycos brand in the US and Japan. (Daum acquired Lycos in July 2004 at US$ 95M). Daum also operates a blog aggregation site called 365bloglink.com in China.

Think of Daum as a Korean Yahoo: Both companies had been market leaders up until some years ago, but now they are chasing the top dog (Naver and Google respectively) and are seemingly more open to the external developer communities. In both companies, people are generally more friendly and less cocky, too. Could be my stereotyping, so take that with a grain of salt.

The company recently held its third company-wide developer conference. The venue was the country that reportedly creates two New York cities a year - China. (The first two Daum dev cons were held in Korea). These days, no one seems to doubt that China is the new US.


Various technology seminars were held in multiple tracks. What caught my attention though was the keynote speech from the Daum founder and Lycos CEO, Mr Jaewoong Lee. Lee said "Attempts to solve problems of human society may from many different angles, but eventually it's technology that fundamentally changes the world. Whether it's global warming, shortage of food and water, education issues in third world, whatever - it's the technology that can address these issues."

Lee also said "But innovation is nothing if there is no underlying value in it - Innovation is actually defined as creating a new value. A company shouldn't focus on business model, brand, marketing, before it does on the value."

I cannot agree more with what Lee said. I mean, it all sounds so cliche, but the harder you try to develop a new service, the more you realize it all really comes down to whether or not you are creating a new, meaningful value. If you are not convinced you're creating a significant value for the customers, you really shouldn't roll out that new service. Because doing so would be to develop a new service for the sake of developing a new service.

Korea's Naver is now the world's 5th search service provider, behind Google, Yahoo, Baidu, and Microsoft.

According to comScore's qSearch 2.0 service, more than 37 billion searches worldwide went through Google in August. That's about 60 percent of all searches, higher than Google's 50 percent in the United States.

Yahoo Inc. was second worldwide with 8.5 billion, followed by Baidu at 3.3 billion, Microsoft Corp. at 2.2 billion and NHN at 2 billion.

Fittingly, NHN's stock price is soaring (NHN is the company behind Naver - NHN's other important business is Hangame, the gaming portal).

NHN stock price over the last year

Meanwhile, according to Korea's eDaily, Naver denies the widespread rumor that the company runs a team in China to "manufacture" search data and hand-pick quality search results. Yeah, right.

But whether Naver has used human labor or not, I believe its position as the world's fifth search service provider is a remarkable feat to achieve. As a startup person, I just hope that the 800-lb gorilla doesn't trounce all startups in Korea and stifle startup innovations. Companies like Nokia have shown that a company can both foster startups and be successful.

Working in the internet industry is certainly filled with hair-pulling stresses, but sometimes good things do happen. Today, I had lunch with Scott McNealy, the Chairman of Sun, along with my partner Chester and 5 other CxOs from other internet companies like Ebay Korea, NCSoft, and Haansoft. I was thrilled - for God's sake, what's the chance of an average Joe like me having lunch with the president of a $14 billion-a-year company?

McNealy visited Korea to deliver a keynote at KES 2007 and presumably to check up on his major clients, but he made time to talk with us, the CxOs of Korea's internet companies, because he believed companies like us were fast becoming his core customers. Sun is shifting its focus from selling-$50,000-apiece-servers business to helping out internet services that deal with lots of user-generated content.

I asked McNealy what's Sun's strategies for maximizing profits from the technologies that Sun invented and then later open-sourced, such as Solaris and Java. Hardware is fast becoming a cheap commodity, which means it's a price-driven market and people might use Sun's open source technologies but not necessarily buy Sun's rather expensive equipment. Almost all the startups I know use LAMP setup (Linux+Apache+MySQL+PHP), and not one of them runs Sun servers on their backend. To this question, McNealy said Sun's hardware business is flourishing with 15+ years of consecutive corporate profits (so in short, "Don't worry about how we make our money, coz it's my problem, and we happen to make gobs of money anyhow"). He also said that sharing a software technology and then tying up hardware with that software technology would be an evil strategy.

After today's meeting, I have great respect for him - instead of fishing in Mediterranean, this billionaire is out here talking to a guy from a 25-person company! Besides, he knew all the details about his company, down to all numbers and figures, and he carried the newest SPARC chips in his pocket. One last thing though: he still doesn't like Bill Gates that much.  

Yesterday I and Chester (my partner) had a lunch with Bernard Moon of Goingon Networks. Turns out, he's a terrific guy. Originally from Chicago, he's got this midwestern friendliness. On top of it, Moon has extensive experiences with technology.

I'll save all the business talks from the lunch. Just one thing to be noted, though: The K-group. There are quite a few Koreans working in the Bay Area, but the industry network among the Korean tech professionals around the Bay Area have been pretty weak. K-group looks to be quite new, and as of August '07 there are around 300 members. Given there are expected to be tens of thousands of Koreans in the Bay Area, many of them presumably working in the tech sector, I guess the K-group has some large room to fill with new members.

I get this question so frequnetly that I consider replacing my personal profile on this blog with the answers for it: Why is Korea so dominated by Naver? Lucas, a terrific guy who recently joined SEOmoz.org and pays regular visits to this humble blog, is the latest person to ask the question.

Without further ado, I'll jump right to the point.

1. So how dominant is Naver in Korean market?

According to a recent study, Naver has 78%+ search market share in Korea, so it's pretty dominating. Naver's Q&A service has 70 million+ entries. I recently saw a news that Wikipedia (English ver.) recently reached 2 million entries. With Naver Q&A being in much shorter form than Wikipedia, no direct match is possible here, but still it's a remarkable feat for Naver.

2. What's their strategy?

These days everyone talks about platform strategy, but another good ol' success strategy in a business is to build a "benevolent cycle". Naver did just that, with the Korean web content - they built a self-reinforcing cycle around Korean web content, all within their walled garden. It's like a giant flywheel now - with so much of momentum built up, even Naver itself seemingly can't stop its growth.

What the heck does that mean?

1) Naver successfully collected virtually all the web content written in Korean out there, and stuffed that content onto its giant DB.

How? Two ways. First, by commercial contracts with various media. A little known secret is that Naver's aggregated news service (since circa 2000) played a critical role in its ascent to a household name from a tiny internet venture. Young people, when they came home from work or school, didn't flip through newspapers as their fathers did - they turned on the computer and browsed through Naver News, which conveniently collected all the news from various news media. Even to day, the #1 time killer on the net in Korea is Cyworld and Naver News. Naver still continues the practice of signing good content prodviders. (a Yahoo strategy, maybe?)

As well as this "B2B aggregation of content", perhaps more importantly, Naver collected content from general users ("B2C aggregation of content") by providing them free, dead-easy content entry systems in Naver blog and Naver Knowledge-iN (Q&A). But often the content wasn't only produced: It was also copied.

2) Content multiplication through copies

Naver didn't actively discourage people from copying others' content. For example, Naver blog UI has a "scrap" button, which allows one-click scrap of someone else's content onto one's own blog. This led to a culture where people think copying someone else's content is a normal thing to do.

As a result of this massive copying, the already humungous Naver DB have swelled even more. But what's the use of massive DB without an effective search?

3) Search, aided by human effort

It's well known that Naver uses human resource to hand-pick good content and provide those content as search results. For example, let's say a national soccer match between Korea and Japan just ended. Naver's part timers quickly collect all the information, such as the video footage of the game, pics, blog entries, relevant newspaper articles, etc, and put them onto Naver DB. When a user enters a query "Korea-Japan soccer game", which is even often given on Naver's front page as the suggested search query to guide the users to their pages, the human-picked content comes out as the top search result, along with other machine-found results. (Both hand-picked content and machine-found content usually come from within the Naver DB; little content indexed from open web is presented.) People say "Wow, Naver does have all information of the world!", and their trust in Naver search becomes more solid.

But the problem of this human-aided search is scalability. Try any "long-tail" search queries, such as Ruby on Rails, and see what results come out on Naver. So Naver apparently focuses on the "head" not "long tail", but it pays off, as most people are interested with the head anyway. I mean, who cares about Ruby on Rails? :)

Anyways, through the cycle 1)-3), namely a "content cycle" that spans content production - content multiplication - content search and consumption, Naver successfully built a "content kingdom" - at least for most popular topics for Koreans.

All this is within Naver's closed walled garden, but when the walled garden is big enough, people are generally content with it. Think the world in the Truman show.

3. Naver Knowledge iN (the Q&A service) - what's the success factor?

One of the most well-known success cases of Naver is its Q&A service called Naver Knowledge-iN. So what's the success factor?

On Naver Q&A service, if you post a question, chances are you will get fairly good answers within a day, or even within hours/minutes. The person who asked the question provides "Knowledge Point" to the person who gave a good anser. (This point system was benchmarked by ).

But the way I see it is this. Naver Knowledge iN was a very smart act. Typical search engines, Google and Yahoo and all, tried so hard to collect all the information residing on the web pages, and analyzed the web pages to give good search results. But then, does all knowledge in the world reside on web pages? Perhaps not. Much of knowledge is still in the people's heads.

How can we take those knowledge out of people's heads? By getting the people with the knowledge to talk about it. And what's the best way to get them to talk? By asking them a question about a topic they are very familiar with. I mean, who would turn back to someone who is asking directions around a place he's familiar with?

4. Any cultural difference behind all this?

Yes and no. Korea is a highly homogenous country where there's a certain dose of what Gen Kanai of Mozilla calls "monoculture". But Naver's success is the result of smart planning and brilliant exeution, and it's not something one can simply call "a Korean thing" and forget about. Put more simply, Naver could have succeeded in other countries, too.  

Orkut can be called a Brazillian thing and forgotten about (no offense to Brazil - I think Brazil has a huge potential and perhaps it's already a big market, I'm just talking about what Google is doing with Orkut), but Korea has the world's 6th largest internet user base, and is the perennial early-adopting market.

Up to some years ago, Yahoo was the number one search service even in Korea. And then it was Daum. How Naver dethroned Yahoo and Daum to become the uncontested leader in the vibrant Korean market is well worth studying.

Bernard Moon, the co-founder of GoingOn Networks, recently contributed a piece on VentureBeat with his views on the future of social networking. It's a good piece - I recommend you give it a read.

Here, I am quickily adding some thoughs on the piece.

1. Facebook will cool down, but won't go away anytime soon either

In Korea, 90%+ people were already using Cyworld even before Myspace and Facebook became widely known in the US. Therefore, looking at how Cyworld is doing now in Korea might give some clue as to how Facebook and Myspace will be doing in some years later.

The answer is, Cyworld is still doing okay. Of course the novelty is long gone, and people don't put up their content as religiously as they did before. Many people leave Cyworld complaining their "Cyworld fatigue". But, one thing to be remembered is, kids are growing - i.e. constant influx of new users.

In Cyworld's case, as I mentioned in my CNN piece, the service sees a constant stream of new users - mainly early-teenage girls. Twelve year old girls would do whatever they think 20 year old girls do, like putting on a makeup, and Cyworld is one of them. I think this will apply to Facebook as well - college is the environment where there's a constant influx of new people. I can't agree more that Facebook won't last forever (nothing does), but I also think it won't disappear overnight eiter.

2. Will social networks last?

I think a longer-lasting concept than social networking is "communication". People have always wanted to stay in touch. Communication between people have always evolved, changing its shape, from fixed line telephone calls to IM to SMS to social networks to content sharing. But at the core of those various types of interactions lie the fundamental needs to communicate.

In other words, when we boil down the human interaction to nodes (people) and lines (communication), the real value might lie in the lines, rather than in the nodes. Traditional thinking from social networking put a lot more values in the nodes, asking people to produce as much content as possible and publicly share them.

But we know people are all too busy for that. The initial interest and fun of producing/sharing content on the web wears off, hence the "fatigue". But people might get tired of producing content, but they will never get tired of communication with other people, and a more efficient means of communication will only help there. So the right question is perhaps, how can social network evolve in a way that can make communication between people more effective?

3. Profile-based recommendation: Is it the future?  

Is life always about personalized recommendations? If I like Red Sox, Jazz, Apple Macbook Pro, should I get recommendations on those stuff for the rest of my life? One of my professors bought a present for on his friend's birthday, and this friend happened to be a gay (the professor is not). As a result, the professor gets recommendations on new products for gays all the time!!
 
I think life consists as much of stumble-upons and serendipities as it does of personalizations and recommendations. I think human life is much more complex than what machines and algorithms can infer, calculate, or let alone predict. IMHO, the only possible way of satisfying the needs for both personalized recommendation and stumble upons (even "pleasantly whacky" suggestions) would be from the friends who know a person well. I think this is where the current social networks isn't capitalizing on the potential effectively, and can therefore evolve into.

A presidential election is certainly a boon for the media industry, in any countries. Elections also often serve as a launch pad for the "new media" - a good example is Ohmynews, which rose to the fame helped by the '02 presidential campaign.

That's why many new media companies in Korea, often called the "UCC (User-created Content) companies", are eyeing squarely for the upcoming presidential election. One such example is the recent "Meet the Bloggers" where a presidential candidate, Mr Kuk-hyun Moon, had an open conversation with 50+ bloggers. (Note: the event was run by our company.)

The whole event was televised real-time via Gom TV, a leading online video service. According to Gom TV, Meet the Bloggers was the first non-entertainment program that made to the top 10 most popular programs list. It will be interesting to see which UCC companies will capitalize on the upcoming national election most successfully.

I and Mr Moon, a Korean presidential candidate

(Photo in courtesy of Bklove)

The Daily Kimchi says my blog is ranked third among blogs on South Korea in terms of "blog juice." Blog juice is calculated by Alexa rank, Blogline subscribers, Technorati rank, and incoming links to the site. Well, this is quite an honor, especially given that I've been quite lazy with blogging these days (Crunch time for work, that is.) I can only say, as the Japanese say, "Isshokenmei Gambarimasu", or I'll do my best.

4.6 The Marmot's Hole 117 / 448598 / 15536 / 4966
3.9 The Metropolitician 43 / 333590 / 31152 / 502
3.7 Web 2.0 Asia 171 / 860236 / 36657 / 206
3.7 Expat Jane 21 / 1,026,144 / 14,052 / 361 *added*
3.6 My Korean Kitchen 89 / 343,464 / 33,650 / 247
3.5 The Daily Kimchi 28 / 399988 / 33440 / 264
3.1 Seoul Daily Photo 5 / 1191008 / 26017 / 443
3.1 North Korea Zone 211 / 2278802 / 110168 / 67
3.1 One Free Korea 31 / 1273190 / 53381 / 638
3.1 DPRK Studies 44 / 1130286 / 70762 / 706